Two suggestions:

- perform your experiment at a primary vacuum level
- model the exact clamping conditions of your device

This should help to converge...

Lionel


Chua Bee Lee wrote:

> Dear All,
>         Thank you for most appreciated response on my last posting on
> "Multiple Resonance Frequencies". I'm now in the process of collating all
> the responses that I have and will be posting the summary shortly.
>
>         At the moment, I'm running into another situation whereby my
> experimental resonance frequencies is consistently lower than the
> analytical and FE simulated ones. I've tested around 20 devices and all of
> them showed likewise. Here are some of the values that I have:
>
>         Analytical : 22.2 kHz
>         FE (PATRAN 2D Shell) : 22.2 kHz (2nd Mode)
>         Experimental : 15.9 kHz
>
> I have the analytical calculation and FE simulated done using E as
> 1.79e11, but by changing E to the smallest sensible value of poly,
> 1.4e11, the natural frequency obtained is still only at 19.6 kHz.
> Can anyone spare any suggestions on that ?
>
>
>         Thank You.
>         Bee Lee CHUA
>         Department of Mechanical and Production Engineering
>         National University of Singapore
>



--
+++++++++++++++++++++++++++++++++

LIONEL BUCHAILLOT
IEMN / Dpt ISEN
Silicon Microsystems
CITE SCIENTIFIQUE
AVENUE POINCARE - BP 69
59652 VILLENEUVE D'ASCQ CEDEX
Phone : +33 3 2019-7838
Fax : +33 3 2019-7884

+++++++++++++++++++++++++++++++++